Estimating Individualized Treatment Rules Using Outcome Weighted Learning
By Yingqi Zhao, Ph.D Major Work: 1. optimal treatment rule $$d^*$$ can be estimated within a weighted classification framework, with weights determined from the clinical otucomes. 2. Replace 0-1 loss with convex surrogate loss with SVM via hinge loss hinge loss function image 3. OWL approach beeter select targeted therapy while making full use of available [...]